Hi Delphinium
This is a major problem with the way in which the 1901 census project was carried out.
The original transcription was done by guests of Her Majesty, most of whom probably had no real interest in what they were doing. This led to large numbers of people called "Ditto" appearing in the transcription! It also meant that if the original handwriting was unclear, the transcriptions were often not accurate.
The transcriptions were then sent to India to be prepared for the internet, where place names and family names were totally unfamiliar.
Hopefully, because they are being done by family historians with an interest in the accuracy of the transcriptions, the FreeCen project should produce better results!
As other people have said, it is vital that we transcribe EXACTLY what was written on the original - there were enough chances for errors to creep in during the original information gathering and recording process without us making alterations 100 years later. We must remember that a large proportion of the people giving the information could not read and write, so the enumerator had to write down what he heard. If the person giving the information had an unfamiliar accent, or a speech defect of some kind, this could lead to all kinds of errors. For example, how many people outside of Kent would know where "Astonlye" is? Or "Uffam"? Or "Walltum"? (answers on a postcard, please!
)
Of course, when we are writing up our own family history, we can "correct the mistakes" by reference to other records, such as birth certificates, baptisms, etc. However, it is still worth noting, maybe as a footnote, what the original record shows.
Regards, Bill